Samsung India's R&D Adds Hindi to Galaxy AI, Boosts Tech for Additional Languages

Samsung's research and development centre in Bengaluru has created a Hindi AI model for Galaxy AI. The technology has also been enhanced for other languages like Thai, Vietnamese, and Indonesian, according to the company.

Representative image

The Samsung R&D Institute India-Bengaluru (SRI-B), the largest research centre outside Korea, collaborated globally to develop AI language models. These models include British, Indian, and Australian English.

Challenges in Developing Hindi AI Model

Creating the Hindi AI model was complex. The team had to cover over 20 regional dialects, tonal inflections, punctuation, and colloquialisms. Additionally, Hindi speakers often mix English words into their conversations.

"To build the speech synthesis element of the AI solution, we carefully reviewed data with native linguists to understand all the unique sounds and created a special set of phenomes to support specific dialects of the language," said SRI-B Head of Language AI Giridhar Jakki.

Data Collection and Collaboration

The Vellore Institute of Technology contributed nearly a million lines of segmented and curated audio data. This data included conversational speech, words, and commands. Data quality was crucial for incorporating Hindi into Galaxy AI.

Working with universities ensured Samsung used high-quality data. "The company said developing the Hindi model for Galaxy AI required the team to carry out multiple rounds of AI model training with a combination of translated and transliterated data," the statement added.

Global Language Support

SRI-B also worked on AI language models for Thai, Vietnamese, and Indonesian. The Galaxy AI platform now supports 16 languages, allowing users to expand their language capabilities even offline.

Hindi's phonetic structure includes retroflex sounds, which are not present in many other languages. This complexity required careful attention during development.

At present, several companies are developing AI models for Indian languages. Hindi is one of the major languages being focused on.

Samsung brands its AI technology platform as Galaxy AI. This platform aims to make language capabilities more accessible globally.

The collaboration between SRI-B and global teams highlights Samsung's commitment to advancing AI technology across various languages.